Image analysis – Image compression or coding – Quantization
Reexamination Certificate
1996-07-02
2001-01-09
Au, Amelia (Department: 2723)
Image analysis
Image compression or coding
Quantization
C358S466000
Reexamination Certificate
active
06173081
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing method and apparatus for quantizing a multi-valued image into a binary image, a four-valued image or the like.
2. Description of the Related Art
Hitherto, an image processing apparatus of the foregoing type, such as, for example, a color ink jet printer, inputs data formed of a multi-valued RGB data or the like via an interface, and converts the input data from RGB values to YMCK values by color conversion processing within the printer. Furthermore, the YMCK multi-valued data is converted into final binary drawing data by a binary process provided within the printer.
Since a color ink jet printer having the above-described construction is a binary display unit which is incapable of controlling the density for each pixel, an error diffusion method is generally widely used as a binarization method in order to reproduce a gradation image.
The first algorithm of the error diffusion method was introduced by Floyd and Steinberg in 1975 (“An Adaptive Algorithm for Spatial Gray Scale” in Society for Information Display 1975 Symposium Digest of Technical Papers, 1975, pp. 36-37). This method globally makes a compensation for a local error by comparing input data with a threshold value for each scanning pixel, and by feeding back an error which arises from the binarized results to the peripheral pixels which have not been processed.
FIG. 12
is a schematic block diagram illustrating the structure of a binarization process by an error diffusion method. As is shown in the figure, an adder
11
inputs an image input signal and a quantization error from an error filter
14
, and outputs an addition result value after both of the input values are added to each other as its internal process (here, for the sake of convenience, the definition areas of the image input signal are assumed to be 0 and 1).
A binarization processor
12
inputs the addition result value, compares the value with a predetermined threshold value (here, assumed to be ½) as its internal process, and outputs, as an image output signal, “1” when the addition result value is greater than the threshold value and “0” when otherwise.
A quantization error calculation means
13
inputs the addition result value and the image output signal, calculates the difference between them as its internal process, and outputs a quantization error amount.
An error filter
14
inputs the quantization error amount, calculates the error amount corresponding to the amount of the addition of the next pixel after the quantization error amount is multiplied by a peripheral distribution ratio of the error amount, and feeds back this amount to the adder
11
.
This error diffusion method, as compared with a dither matrix method, has an excellent feature in that the resolution of a reproduced image is higher.
However, in a binarization method employing an error diffusion method, there are problems: for example, geometrical interference fringes occur in a specific density area due to the influence of the frequency characteristic possessed by the system itself, causing the image quality to deteriorate considerably. This phenomenon is caused by the fact that the error diffusion system is itself formed of a closed loop structure, and this cannot be prevented because the error diffusion method itself is configured to surely distribute an error caused in each pixel to the peripheral pixels (e.g., via a peripheral distribution ratio as discussed above). As a main factor which determines the characteristics of such an error diffusion system, the characteristics of the error filter can be cited. These characteristics are a main factor which determines the pattern in which the geometrical interference fringes occur.
For this reason, some proposals have hitherto been made for improving the error filter. However, no method for providing a radical solution to such problems has yet been found in any of the proposals. For example, other patterns are used to merely change the generation pattern of the interference fringes, or even if random signals are used to erase the interference fringes, the image contains white-noise.
SUMMARY OF THE INVENTION
The present invention eliminates the drawbacks of the above-described related art. It is an object of the present invention to provide an image processing method and apparatus capable of obtaining a high-quality image by an easy processing method.
It is another object of the present invention to provide an image processing apparatus which is capable of effectively suppressing geometrical interference fringes which occur by a conventional error diffusion method by periodically varying an error amount to be diffused within a range of values up to and including the error amount which is generated by a quantization process in an error diffusion process.
To achieve the above-described objects, there is provided an image processing apparatus, comprising: input means for inputting multi-valued image data; quantization means for quantizing the input image data into output image data whose number of levels is smaller than that of the input image data; and correction means for correcting a quantization error which occurs during the quantization process by the quantization means by diffusing the error into image data which have not been quantized, wherein the correction means periodically varies the error amount to be diffused within a range of values up to and including the error amount which has occurred by the quantization process.
It is a further object of the present invention to provide an image processing method and apparatus which is capable of achieving uniform formation of dots which are formed by error diffusion, by periodically varying the error amount which occurs locally by quantization by means of a feedback amount control section and by shifting the phase of the error amount by a half-period, and which is thus capable of effectively suppressing an occurrence of geometrical interference fringes.
It is a still further object of the present invention to provide an image processing method and apparatus which is capable of uniformly diffusing an error by periodically varying an error amount which occurs locally by quantization by means of a feedback amount control section and which is thus capable of effectively suppressing an occurrence of geometrical interference fringes, and which is capable of forming a more natural image by suppressing an influence by the period signal itself upon the image by giving the image an overall noise appearance in accordance with a disturbance signal.
In view of the foregoing, in one aspect, the present invention relates to an image processing apparatus comprising input means for inputting multi-valued input image data, quantization means for quantizing the input image data input by the input means into output image data having a number of levels which is smaller than that of the input image data, and correction means for correcting a quantization error which occurs during a quantization process of the quantization means, the correction means correcting the quantization error by diffusing an error amount into input image data which have not yet been quantized by the quantization means, wherein the correction means periodically varies the error amount to be diffused into the input image data within a range of values up to and including the quantization error.
In another aspect, the present invention relates to an image processing method comprising an input step of inputting multi-valued input image data, a quantization step of quantizing the input image data input in the input step into output image data having a number of levels which is smaller than that of the input image data, and a correction step of correcting a quantization error which occurs during the quantization step, the correction step correcting the quantization error by diffusing the quantization error into input image data which have not yet been quantized, wherein the correction step periodically varies an error amou
Ahmed Samir
Au Amelia
Canon Kabushiki Kaisha
Fitzpatrick ,Cella, Harper & Scinto
LandOfFree
Image processing method and apparatus with variation of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method and apparatus with variation of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method and apparatus with variation of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2548649